A more successful landing than this.
I thought I'd check how often we'd used that title here: the results are amusing.
It could be a fake, but a quick google reveals lots of people taking it fairly seriously, and no one I can find saying "Clearly a fake and here's how I know."
(Oh, and I have to change the post title now. For future reference, it used to be "I, for one, welcome our new avian overlords.")
Clearly a fake. Probably made by the company that sells the parasail it's allegedly made from (you can see their logo pretty regularly).
The Gizmodo post originally had updates with some comments that made very salient points; why does everybody run away like that when he's taking off? Why is there always a moment in the videos where you see the apparatus "working" where he's either offscreen or occluded by people? Why do his wings all fly out behind him as he's climbing?
I would add to those: why are both of the videos where they show the mechanism filmed with tons of lens flare that washes out detail? Why aren't there any videos of the mechanism working on the ground (except that ricidulous one of something with no fabric attached)? Why do they stay so far away as he lands (yet they still can't hold the camera steady)?
But more than that, his body in the "flying" videos doesn't look real at all. It read as so obviously, immediately fake to me when I saw it yesterday that I'm actually sort of surprised that everybody hasn't already concluded that.
6: I have to change the post title now
?? Are the archives the boss of you?
I mean I feel a little bit like I'm being all "YOU CAN TELL CUZ OF THE PIXELS" but I am perfectly confident I'm right so that's okay.
Why do his wings legs all fly out behind him as he's climbing?
(sorry)
Well, now I'm sad -- this isn't the kind of thing that I have a lot of judgment about. I did have a moment of "the proportions look all wrong -- his body/wing ratio looks hawklike, and at human size, shouldn't he need much more wing?" but if you're not credulous sometimes, where's the fun in life.
6: No, but this is clearly a better title.
Sure, it *looks* fun, but the preflight surgery to hollow out your bones hurts like hell.
Looks fake to me, too. As LB says, the wing-to-weight ratio looks wrong. As per Tweety, the body shape is really odd at one point -- he has his legs hanging down in a drag-tastic fashion, at one point -- and the fabric on the 'wings' looks too flappy to be producing lift. And where the power? Whatever he has on his back can't be big enough to produce lift fora 150-200lb human.
To defend the legs, if it were real, that didn't look implausible to me. He was running, then took off, then after a few seconds lifted his legs out straight behind him. How else would the legs look? (But I really can't judge the plausibility generally myself. I'm only believing it because grounds for disbelieving it seem as though they would be so obvious for someone with expertise that no one would have taken it seriously at all. The fact that there's an argument, rather than everyone reputable saying "Impossible, and here's why" suggests to me that maybe there's something.)
I had an initial reaction that there was no way that dinky wing twitching was going to produce significant lift, but being a sucker for video evidence I dismissed it.
No, but this is clearly a better title.
OK. I did some tests and you are right. 7.9 rat orgasms better to be precise.
Looks legit to me. I don't think this is somehow outside the bounds of the current state of human technology.
I don't think the wing-to-weight ratio is as off as people are saying. The wings look somewhat smaller than a hang-glider to me, which makes sense, because they are creating additional lift through flapping, wearas the hang-glider just generates lift by moving through the air.
He drops his legs when hes climbing at the beginning because that's what birds do.
On the other hand, its on the internet, which does make it highly suspicious.
16: You only keep getting approval because the IACUC is packed with rats, you know.
Def. fake. If you look at the earlier video, you can see he's flapping the wings himself.
http://www.youtube.com/watch?v=Q0tKFOcHyrI&t=1m40s
As per another blog post elsewhere on this, the camera pans down to the ground [no good reason], there's a clear edit, and when it comes back up, the right wing is different, you can see where the cord is attached.
Is Martin around? Any local Dutch-language coverage? This was supposed to be in a park, if it's real there should have been a whole lot of "WTF!? Freaky birdman overhead!" on twitter.
re: 17
There's no way he has the speed to lift himself with the drag from the legs.
I hate to be internet-sceptic guy [as they are usually arseholes] but it is pretty clearly bullshit.
I thought he was supposed to be flapping the wings himself with powered assist. You can certainly see strings.
14: holding your legs out straight behind you is extremely tiring. Hang gliders have a bag they stick their legs into that is strapped to the glider. Anyhow, you need a ton of leverage to hold your body that way; it would naturally tend to tip you back, which would be terrible for gaining lift.
I still find it comically unbelievable, sorry.
re: 23
Yeah, but there's a tension between the drag, and the need for lift/leverage and so on. I'm not seeing the power source to overcome that.
At this point, I'm convinced by the skeptics. But I'm still hoping they're wrong.
One more for the skeptics; I was really surprised, even watching credulously, at the controlled nature of the landing. No staggering or face-planting, he touched down like a gull. That's weird for his first flight ever.
There are so many hinky things about the filming -- each vid has a pan or cut away right before the key moments -- even if there was more evidence for the tech behind the flight. N.B. he doesn't claim he uses cables to pull the wings, but that he uses wii-remotes and bluetooth, iirc.
When birds generate lift there is a big asymmetry between the up and down strokes; they fold their wings when lifting them to avoid pushing up against the air. This video shows wings that just twitch up and down, which might make the contraption bounce, but can't possibly lift sustainably.
I mean compare with this. Winged, heavier than air flapping flight is an extremely difficult engineering problem.
Self-powered flight is really heard. Birds have basically no legs and weigh next to nothing under all those feathers. Even the largest flying birds weigh less than 35lbs.
Only had time to take a quick look at the video itself (and then ran off to find the Brewster McCloud clip because pop cultural references are more important than the truth), but absent a slew of ancillary data and confirming information I'd be skeptical. I did/saw this great thing and all I got was this lousy t-shirt video as evidence is the world of Woolly Mammoths crossing a river in Siberia, Ivory-Billed Woodpeckers in Arkansas and Cold Fusion.
This reminds me of the liquid mountain walking on water one.
Bird wings also scale allometrically, so a bird the size of a human would have wings that were proportionately much, much larger than the wings of even the largest birds. And a bird the weight of a human, far, far larger than that.
Everyone should watch the video in 29.
The inevitable mythbusters debunking will be fun, in any case.
Has this one been Zaprudered to death yet? I need closure.
Has anyone seen the show where a bunch of pseudo-scientific Bigfoot hunters go around and search for Bigfoot? It's so great, they'll do things like "debunk" Bigfoot videos from yokels ("we know that Sasquatches don't nest in this kind of forest cover, so I'm sorry Jethro, but you seem to have been had") and then go out into the woods with infrared sensors and come up with the most awesomely fake science explanations as to why THEY are on the trail of the REAL Bigfoot. "We've got a response knock in the appropriate aural range. A RESPONSE KNOCK IN RANGE PEOPLE!". Anyway, it's a great show, you all should watch.
Wow. Guess I bought that Kansas beachfront property. I assumed that the mechanism was providing the extra power, although couldn't figure out the math. Thanks for the roller coaster start to the morning. I need to go watch some cute kitten videos now. Those aren't fake, right? Right?!?
I guess it just don't think it would be physically impossible for someone could fit together a flying contraption with a couple high-torque servo-motors, a large, high current lithium-ion battery, some blue-tooth shit, a circuit board, some carbon fiber rods, and a few square yards of rip-stop nylon. Why have the technology, and its not stupid expensive. Why wouldn't some dude be able to pull this off?
This thread is a great example of etiology of Unfogged Impostor Syndrome (i.e. the idea, which I'm told is not uncommon, that everyone else here knows ALL THE THINGS, a hundred times what one knows, oneself.) I mean, you people know strangely much about birds.
Has this one been Zaprudered to death yet?
This is funny partly because Wired reported is as real yesterday and partly because it's a true Zapruder analysis, in that it's almost all about the video.
Has this one been Zaprudered to death yet?
This is funny partly because Wired reported is as real yesterday and partly because it's a true Zapruder analysis, in that it's almost all about the video.
you people know strangely much about birds
Thanks to urple.
The mythbusters dude buys it.
Anyhow, to 40, I don't know that I think it would be impossible to build a flapping glider that would work a bit better than a non-flapping glider. I would just point out that, in this case, totes fake.
There's no way he has the speed to lift himself with the drag from the legs.
It's not a fixed-wing aircraft - what matters is the relative airflow over the wings as they move, not the whole contraption's air speed. (A helicopter has no trouble moving slowly or hovering; its rotary wing has to keep moving.)
Its not like birds need to get a running start. Well, except for loons.
34, 28: Totally correct. The asymmetry of the wing beat is key to how birds fly. It's not just straight up and down.
And the wings are way too small and not variable enough in shape. Those wings are similar in shape to albatrosses - narrow - but not nearly long enough to be proper gliding wings in the albatross style. Other soaring birds, like hawks, have wide wings with long primary feathers but obviously these wings are not shaped like that either. Small flappy birds like sparrows have almost triangular-shaped wings - long along the primary edge - but this shape is modified during different parts of flight. When birds take off, there is a lot more surface area compared to when they are flying level at a constant speed.
Finally, the flapping speed is too small and the flap (I don't know what this is called) distance? The distance the wing travels between the top and bottom of the flap is laughable small for flight. Albatrosses, for example, look like the tips of their wings move over half their body length. Not to mention the curve in the wings.
Fake.
I'm confused about how the bluetooth and wii controllers play into all of this. The motors give more force when he flaps his arms. Do the wii controllers tell the motors where his arms are? Does he have to exert any particular force with is arms, or to the wings just mimic the way he holds his arms?
Has anyone watched much of the other 13 videos? Potentially would removes some of my concerns voiced in 32.
El Reg thinks probably not real, but clearly trying to keep their options open.
51: there's only one other video of the built wings in action and it has similar problems to the one LB linked (worse, actually; they do the whole "oh the camera was pointed at the ground in our excitement" thing).
On the other hand, the dude's twitter feed goes back to July (and sort of eases into the project somewhat naturalistically) and they clearly actually did build something, so there was certainly a fair bit of time/effort put into this.
A lot of marine or water-associated birds run to takeoff. Ducks and geese and loons from the water surface (loons sometimes get stuck on small lakes because there isn't enough room to get their running starting for flight). Seabirds like puffins and murres run on land and water to take off. Or just jump off cliffs to get airborne.
I was thinking an albatross would be the best model for human flight using flapping wings, but now I'm going to go with loons. Although you'd have to work on getting your legs to stay out of the way and not dangle down. Loons can't really walk on land because their legs are shifted so far back on their body.
I'm betting the motors do something useful--e.g., once the thing has been pulled into the air by whatever left the tire marks on the grass, the motors helpfully flap the arms of the dummy they attached to the wing.
And like Great Albatrosses, Dutch engineering students weigh up to 25lbs.
51: the pedal-powered ornithopter they link to is both much cooler than this and illustrative of some of the things wrong with this video; for one thing, it has a wingspan more than 4 times longer than fakey fake's wings.
The Mythbusters guy makes the point that with a large enough headwind and a running start it's possible to do something close to this with a fixed glider, but I can't believe the flapping is adding anything. I repeat my call to watch Tweety's video and compare the complexity of that motion. Also, there's a helium jellyfish robot, so.
loons sometimes get stuck on small lakes
Now I'm sad.
It's a pity that Lunchy has taken Flip from us, as it seems like he could have something to say about the lying Dutchmen.
56: I was just about to post that link.
58: I know, right? When I first heard that here (I think from Emerson), I wanted to start a loon rescue patrol. We'd go from pond to pond with butterfly nets and small catapults.
The Mythbusters guy makes the point that with a large enough headwind and a running start it's possible to do something close to this with a fixed glider
Down the side of a hill in a strong wind, maybe. The guy's on basically flat ground on a calm day in Holland!
re: 47
Well sure, but it's not really flapping much either, or doing much of fuck all. It's a wobbly hanglider.
they clearly actually did build something, so there was certainly a fair bit of time/effort put into this.
This is why I lean towards it being real. If you are going to put that much time into the thing, why waste your effort on a hoax instead of on building the real thing?
This isn't like the walking-on-water hoax video, where a few mooks spend the afternoon fucking around with some plexiglas at a lake. This would have taken significant investment and dedication.
I wonder if this guy has a track record that could be looked into. Has he created anything else particularly cool in the past? Or has he been involved in hoaxes before?
This would have taken significant investment and dedication.
I mean, it's not as if anyone in the past has gone to an inordinate amount of well-planned trouble to perpetrate a hoax, now is it?
If you are going to put that much time into the thing, why waste your effort on a hoax instead of on building the real thing?
What real thing? A somewhat flappy hang-glider, or a set of wings with which you can take off and land like a bird on a calm day? Because if it's the latter, it would take a whole hell of a lot more work than faking this did (assuming it's even possible to do, which it might be at some point, maybe).
I mean, it's not as if anyone in the past has gone to an inordinate amount of well-planned trouble to perpetrate a hoax, now is it?
Of course not. But I don't think building the real thing would have been that much more difficult. And if the tasks are of the same magnitude, why bother with the hoax?
assuming it's even possible to do, which it might be at some point, maybe
I guess my thought is that, technologically, we have finally reached some point. All the stuff to do this is currently available off the shelf, its just a matter of getting the design right.
We live in the future, man, all sorts of crazy shit is possible.
I don't think building the real thing would have been that much more difficult
The tasks are not of the same magnitude. Building a set of wings that a grown man can flap himself eighty feet into the air with after a thirty-step run-up is going to be a lot harder than faking a film of such a thing.
| part 14/14.
That's the bit I like the best, really. He must be a really good planner.
We live in the future, man
And yet every day, loons go unrescued on small lakes. I demand a national initiative for a fleet of autonomous loon rescue robots. Decorated in soothing, loon-friendly colors and textures, to minimize their anxiety.
Building a set of wings that a grown man can flap himself eighty feet into the air with after a thirty-step run-up is going to be a lot harder than faking a film of such a thing.
I guess I am assuming that he had a significantly greater amount of assistance from the motors than you are. It seemed to me that his arms were mostly used for direction and control.
The video in 29 is awesome.
It's interesting that the more authoritative, bigger sites -- the Mythbuster guy, Wired, Gizmodo -- seem to be leaning slightly towards "real," based on the theory that it could be real; the site LB links to above seems to be from someone who thinks the thing certainly could exist.
I have no idea and don't have enough knowledge to form an idea; if I had to bet, I'd bet fake (or, "flappy hang glider not powered by wing motion"), but real would sure be nice. With next to no relevant knowledge, I'm not seeing why a failure to precisely mimic actual bird flight would be decisive -- the thing (if real) doesn't stay aloft very long or with much control, unlike what birds do. But what the hell do I know.
71: Like an ice cream truck, you could play their crazy laughs as you approach the small lakes so they know help is on the way.
Each time I look at the video it looks faker.
When people say "fake" do they mean "doesn't work nearly as awesomely as the video makes it seem" or "they used camera tricks and a lightweight dummy"?
If the former, how much of an improvement over a regular hang glider does a flappy hang glider have to be before it counts as actually something really cool?
77.1: the latter. They used camera tricks, CGI and a lightweight dummy.
77.2: if somebody could take off and climb from level ground on a still day that would be pretty mind-blowing.
Down the side of a hill in a strong wind, maybe
While I would defer to more expert opinion (I believe we have an expert here), from what I've seen, running or skiing down a slope is exactly how people do take off with hang gliders or paragliders. They often have cliffs at the end, but they're generally airborne before they go over. And the slopes aren't necessarily that steep, nor are the headwinds necessarily that strong. On the other hand, taking off in those conditions... doesn't seem likely, and the video seems implausible to me.
This guy who is a physicist used some video analysis tools and doesn't think its fake or was using camera tricks. I'm not endorsing that analysis, but there it is.
80: That was really unconvincing. For instance, on the hand held issue, some possibilities he neglects: they have a more realistic hand motion simulator than the subject of his previous analysis, they used motion tracking software similar to his to stabilize any effects they added, the video is the product of a sequence of cuts.
I read that, and it's not all that strong -- all he's saying is that the camera shake is real rather than added afterwards, and that the speeds and so on make it not wildly impossible. (I have a hard time believing the chart that says the wings are big enough.)
It's annoying not speaking Dutch: I'm 90% convinced it's a fake now, but what would convince me one way or the other is the presence or absence of local news coverage or local witnesses not affiliated with the project. Someone local who got on the phone with the guy and asked for witnesses would convince me.
all he's saying is that the camera shake is real rather than added afterwards
But all he's shown is that he can't prove the camera shake is fake.
70: | part 14/14.
That's the bit I like the best, really. He must be a really good planner.
It is possible to retroactively change the title of YouTube videos. Some of his earlier videos got noticed and embedded in November/December that are now in the sequence, if they were 7/14 then I'd agree, rather than a retroactive attempt to sequence (which seems to be incomplete (or incompletely labeled at least) in YouTube at the moment).
re:78.1
Yeah, the legs on the 'person' flying the thing are really un-natural looking from some angles. Almost 2D.
Assume no comma splice in 84. Same for all comments retroactively and in the future.
Anyhow, if it's a fake, which it probably is, it seems to be a pretty good one-- that is, it took something that is just on the edge of the technologically possible, set themselves up with cutting-edge equipment that plausibly could have achieved the thing, and then did a video job good enough to fool experts or at least to be difficult to clearly and decisively refute. Even as a fake it's pretty interesting.
Here is a Google Translate translation of a piece by a Dutch journalist(? not sure) who had interviewed him earlier and who tried to follow-up afterward. One damning bit of evidence, the companies he had apparently said he had worked for on his LinkedIn profile had not heard of him.
We try Jarno Smeets several days to reach for a reaction, but in vain. Even the repeated attempts to contact the Dutch professional failed.
This article at BNR Newsradio is blocked for me, but may be worth a read. From the summary that I could translate: Smeets Hagenaar Jarno this weekend would have flown like a bird. His video will be more than one million views. But the story stinks.
Speaking of fakes, doesn't the breitbart.com interview of "Bono" open up all sorts of possibilities? For what used to quaintly be known as ratfucking?
For reference, the first hang glider I could find had 20 square meteres of wing, this dude's contraption has 17.
It may well be a hoax, but I'm not buying "that's physically impossible" as a convincing argument that it is.
87: There are several updates that seem pretty definitive. Don't know if they were there when you read the article.
Nope, they weren't. I still wouldn't take it as "definitive" without sourcing, but the evidence is definitely tending towards fake.
93: I felt like they found some artifacts that said could be CGI, but were less than convincing that the artifacts could only be CGI. Also, the examples they picked were when the aircraft was on the ground, which doesn't make a lot of sense to me... why would the portions of the video on the ground be faked? Those are exactly the parts that wouldn't need CGI.
Most importantly, I also find the use of German in the title of this post pretentious and unnecessary. "The Flying Dutchman" is a perfectly good English phrase and reference and should have been used instead. Wikipedia suggests that the phrase was coined in English before it was used in German.
Once we run out of things to attack in the post, we will attack the poster!
I don't have any particular expertise, but the video sure looks fake.
A fury of meaningless criticism and contempt gives life to the blog and death to our souls.
I kind of feel this post -- something awesome that turns out to be faked -- is truly in the spirit of Ogged. I should post neat stuff I find out about late at night after a hard day more often.
96: Or it could have been "De Vliegende Hollander". But I for one welcome our newly pretentious front-page posters.
I will say that what got me the last ten percent of the way was 89.
100: Yes, more System 1 posts made during periods of ego depletion.
truly in the spirit of Ogged
I had this very thought a few minutes ago.
is truly in the spirit of Ogged
Except for the part where you sexually objectify the dude and then spend half the day defending yourself.
Will no one only LB think of the comment volume?
It isn't too late, LB! Did anything about the dude make you feel tingly?
What percentage of Unfogged commenters have masturbated a hang-glider?
i love that Gizmodo in 87 is all like "haha stupid internet getting all het up" when they were the ones all "IS IT FAKE? IS IT REAL? RANDOM PEOPLE WEIGH IN!" yesterday.
Anyhow, as I have said consistently, and because I said so consistently specifically so I could have this moment: sooooooo fake!
96: Sure, but LB specifically wanted to reference the Wagner opera. Because RACIST.
What percentage of Unfogged commenters have masturbated a hang-glider?
The person, or the piece of equipment?
108 Canine or homo sapiens hang gliders?
I can also tell you specifically how it was faked and where the seams are, if you'd like.
1. They film real dude from close-up.
2. They film real dude in harness as they run randomly away from thing
3. dude randomly runs in front of camera to cover cut to CG
4. CG dude and flappy thing runs and takes off
5. cut to airborne glider (pulled aloft by the vehicle which left tire tracks) with motors and dummy with arms attached to wings.
6. show gopro footage from dummy head on landing approach
7. cut to CGI landing until
8. dude randomly runs right in front of the camera again, leading to a cut to
9. real dude attached to armature kind of jogging, allegedly to a stop.
108: I think you need to ask flippanter.
111, 112: We'll be inclusive of any and all forms of hang-glider. So I take it that gets us up to 2.
What percentage of Unfogged commenters have masturbated a hang-glider?
Assuming "hang-glider" includes "hang glider enthusiasts" we can say with 99% certainty that there is at least 1 such commenter.
God damn it I was pwnd. I guess we know that there is one such commenter and also a high probability that a favorite topic of conversation here has done the same. Now I have grossed myself out and am probably pwnd as well.
So 3. Tell us more, Jackmormon.
Wait, how do you get to two?
I don't understand why Sifu is so skeptical. Didn't he hear the soaring music?
Oh, huh, the CGI people disagree with me on 113. It's real!
122: Of course it's real. Look at how the guy's panting at the end.*
*Unless that's from the masturbation.
120: I reasonably assumed 111 and 112 were asking for clarification based on personal experience. By making the criteria expansive, I can mark them as a "yes" and spare them the embarrassment of a long-drawn out certification process involving CGI experts and whatnot.
I reasonably assumed 111 and 112 were asking for clarification based on personal experience.
Not really a reasonable assumption around here, as a general matter.
I haven't figured out yet what a metaphorical hang glider is yet.
But it sounds like it could be dirty.
I haven't figured out yet what a metaphorical hang glider is yet.
It starts out like doggy-style, but then you jump off a cliff.
It's a metaphorical hang glider but a literal cliff.
I'm commenting from 30,000 feet!
This seemed like the appropriate thread for that announcement. In flight wifi sort of snuck up on me. How long has it been commonly available?
I like how essear is progressively transforming into the protagonist of Up In The Air.
132: How do you type and flap your arms at the same time?
I hope this guy has a recent save point; Wiimotes aren't that reliable.
@137
So that's why the comments have been acting weird?
Also, Crooked Timber seems to have lost the entire month of March, but I doubt that's related to in flight wifi weirdness.
134: I was disturbed to realize I've flown on average 380 miles a day so far in 2012. (The number should start falling rapidly now, I hope.)
Off-topic: a friend/colleague who is also getting bored with our day job is trying to convince me that it would be fun and useful (for us, if no one else) to spend some time on this. I'm kind of semi-on-board with the idea of learning some stuff about statistics and whatnot for fun, and with picking a concrete problem to tackle as a way to motivate learning stuff instead of just picking up books at random and reading, but I wonder: is this actually a worthwhile thing to target? Or is this somehow, in the end, aimed at helping insurance companies to be more evil?
Once known, health care providers can develop new care plans and strategies to reach patients before emergencies occur, thereby reducing the number of unnecessary hospitalizations.
I vote evil.
No, not evil, except to the extent that a good result may bolster Heritage's reputation. It's feeding off a recent New Yorker article on high-utilizers, and seems to me something that would be equally useful in a single-payer system or our current system. A lot of the resulting interventions will be things like arranging doctor visits and connecting to social services.
I wonder: is this actually a worthwhile thing to target? Or is this somehow, in the end, aimed at helping insurance companies to be more evil?
Eh, how likely is it that you will make significant contributions to solving the problem? I'm not trying to be cynical just noting that if you use it to learn stats and end up arriving at the 10th best wrong solution you will have accomplished all of your personal goals (other than winning the $3M) and won't have changed the insurance industry at all.
How likely is it that one can identify patients who are likely to soon make a trip to the ER that could be avoided with cheaper outpatient intervention with any specificity? By far the most robust result of any algorithm will be to give insurance companies a better estimate on how much patients are likely to cost them.
You guys missed the part where, after you've identified likely overusers, you track them down and kill them before they reach the specialist. If we all do our part, health costs will take a serious nosedive.
How likely is it that one can identify patients who are likely to soon make a trip to the ER that could be avoided with cheaper outpatient intervention with any specificity?
No idea! But they're so high-cost, isn't it worth trying to find out?
By far the most robust result of any algorithm will be to give insurance companies a better estimate on how much patients are likely to cost them.
Uh, they've been pouring resources into this very question for ages, and will continue to. Benefit to them: probably marginal unless something no one could have predicted is found. Also, so many of the high-utilizers are poor that there's a market failure in analyzing them. Results here should redound more to the benefit of public programs than to private, or at least equally.
Granted, they don't say whose data it is - maybe insurers' only, since one field is "Member ID." But it should still be helpful to public programs to the extent that the results are made public, whereas insurers' own analyses are trade secrets.
Finally, if health reform goes forward, with community rating and guaranteed issue, even for private insurers the application of the results would be to improve care for enrollees rather than to cherrypick and reciss them.
How likely is it that one can identify patients who are likely to soon make a trip to the ER that could be avoided with cheaper outpatient intervention with any specificity?
Isn't this like every patient who has something bad happen to them on a weekend? I've had to go to an ER for relatively minor things just because it was worrying enough to be worrying and it was a Saturday (most recently, I avoided a hospital but went to this ER place, which was decorated like the nightclub at a W hotel, complete with club music and colored lighting.)
Did they serve drinks?
Lucille Bluth: I'll be in the hospital bar.
Michael Bluth: Uhh, you know, there isn't a hospital bar, mother.
Lucille Bluth: Well, this is why people hate hospitals.
http://www.imdb.com/title/tt0367279/quotes?qt=qt0386484
No, but there was a gym to exercise in while you waited (n.b., you are waiting for an Emergency Room) and a weight-loss program you could sign up for. It was like they were trying to live up to some New Yorker's stereotype of Beverly Hills.
When I was an IT dude for a health insurer, one of our big-ass projects involved figuring out how to incentivise doctors to provide better preventative care to the highest-cost patients.
Basically, the doctor was supposed to call up the patient and say, "so, your insurance company tells me you are high-risk... come in for medical care so I can collect a bonus."
Which was in many ways a good thing to do - it would help the patients and save the company a ton of money - except it also involved massive HIPAA privacy violations in the process of identifying those high cost patients to the various medical practitioners involved.
140: Is the worry that they will rescind/cancel policies for high-risk patients?
Because right now there is a large class of people for whom insurance is overpriced because of adverse selection, and another large class deemed "uninsurable." So even if they make decisions based on this, it should have the effect of making it easier for non-high-risk people to buy health insurance.
And if the ACA remains law, it will be somewhat more difficult to use this to deny care, as opposed to investing in prevention.
OTOH the loss ratio requirements may make it difficult to implement robust preventive care practices, depending on how much can be automated, and how cheaply.
Ah, here's what I was talking about... a PCMH - Patient Centered Medical Home. The idea is to identify the sickest patients, and then steer them into one of these thingies.
151.1: More or less. But I think 143 gets it right; it's not like I would be contributing anything very useful to anyone's evil plans, and I might learn something.
154: I might learn something.
There is the danger that what you might learn would get you hired by the health insurance industry.
More seriously, when this challenge first came out I had a plan for how to attack it with a genetic algorithm that I would brute-force the hell out of using an array of high-end, massively parallelized multi-core graphics cards. But I figured everybody and his brother would be trying that same solution, and probably do a better job at it than me because I'm not actually an expert in any of that stuff. And I'm also lazy.
Did you have a similar approach to the problem in mind, or something completely different?
You're going to best at that with a model that embeds some expert knowledge over a brute force approach, I bet.
Maybe integrate some outside predictors. Is that allowed?
You're going to best at that with a model that embeds some expert knowledge over a brute force approach, I bet.
Well, maybe, but I have know expert knowledge, and the brute force approach would allow me to put one of these bitchez to good use.
You wouldn't use a bitcoin rig for bitcoin? I mean, wouldn't do a damn thing with bitcoin, but it's right there in the name.
No, I think bitcoin is stupid and probably not worth the electricity. But those dudes have done some sweet innovation with teh hardware, and I'd love to figure out how to apply one of those rigs to solve problems in some other, non-bitcoin-related domain.
Running through a bajillion generations of a genetic algorithm on health outcomes seems like it might be one potential application.
Flying elephants were totally possible using Victorian technology.
162.1: they have? Like what? (I haven't followed that world except peripherally, although the thing about the Highlander constant was rad.)
But an elephant wearing a top hat? That's a Regency style.
essear should just be careful not to go too far.
Like what?
Like figuring out you could stick a half-dozen 3200 core GPUs onto a mother-board and create an ungodly amount of parallel computing power in your living room.
169: aside from the "living room" thing isn't that what university clusters have been doing for years now?
But those dudes have done some sweet innovation with teh hardware
Speaking of which: "[The Pirate Bay] plans to move its front-end proxy servers into the sky, creating a network of small mobile computers that are tethered to GPS-enabled aerial drones. The airborne computers, called Low Orbit Server Stations (LOSS), will supposedly be harder for law enforcement agencies to terminate."
169: Yeah but it's a flying living room.
isn't that what university clusters have been doing for years now?
Perhaps, but universities doing it and some Joe with $2500 in his pocket doing it are two different things.
I would have to think about whether I would run into any problems playing with algorithms for this competition on the H/rv/rd Od/ss/y cluster. Probably no one would notice or care?
Anyway, as a general rule, I have no interest in brute forcing a solution to any problem. (Rant about what most people consider good coll/d/r ph/s/cs expunged.)
And if I do decide to play with this, I'd probably be looking for more team members, if anyone else thinks it sounds like fun... (I would definitely need to cap the numbers of hours per week I could think about it, though. The guy who's trying to get me into it is more interested in leaving the field in the near-ish future and could probably invest arbitrary amounts of time, though.)
I used to be philosophically opposed to brute force as well, but I think the advent of cheap parallelism on GPUs has really opened up the possibilities for its use. Computations that used to take years can now take days or hours on desktop hardware. That's just too cool a possibility not to take advantage of.
Anyway, I think it sounds like a pretty interesting problem, but I've got way too many projects on the burner these days to take up a new one. Right now the priority is on building HTML5 games with my kid.
"[The Pirate Bay] plans to move its front-end proxy servers into the sky, creating a network of small mobile computers that are tethered to GPS-enabled aerial drones. The airborne computers, called Low Orbit Server Stations (LOSS), will supposedly be harder for law enforcement agencies to terminate."
LOSS? Really? Come on, guys, it's a network, in the sky. There's only one thing you can call it.
Well THAT sucks. NMM to flying like a bird.
Float like a dinosaur, sting like a con.